In the realm of modern law enforcement and criminal investigation, the use of artificial intelligence in forensic sketching represents a milestone. This project showcases an AI-Powered Forensic Face Sketch and Recognition System intended to aid in suspect identification from eyewitness accounts. The system utilizes a multi-phase method that starts with natural language understanding (NLP) to analyze speech or written descriptions and translate them into rough sketches where these sketches are enhanced to produce realistic facial images. Additionally, the system includes facial recognition systems to compare the created faces against a criminal registry and initiate real-time alerts. This blended solution not only improves the accuracy and speed of suspect identification but provides a significant forensic tool. This also accepts voice or text, extracts feature, creates 2D and 3D face models, and compares them with criminal records contributing to a more comprehensive and interactive profiling system.
Harshitha et al. (Tue,) studied this question.